merge from dev

cmake/mimalloc-config-version.cmake

@@ -1,5 +1,5 @@
 set(mi_version_major 1)
-set(mi_version_minor 0)
+set(mi_version_minor 1)
 set(mi_version ${mi_version_major}.${mi_version_minor})
 
 set(PACKAGE_VERSION ${mi_version})

include/mimalloc.h

@@ -8,7 +8,7 @@ terms of the MIT license. A copy of the license can be found in the file
 #ifndef MIMALLOC_H
 #define MIMALLOC_H
 
-#define MI_MALLOC_VERSION 100   // major + 2 digits minor
+#define MI_MALLOC_VERSION 110   // major + 2 digits minor
 
 // ------------------------------------------------------
 // Compiler specific attributes

readme.md

@@ -20,7 +20,7 @@ without code changes, for example, on dynamically linked ELF-based systems (Linu
 
 Notable aspects of the design include:
 
-- __small and consistent__: the library is less than 3500 LOC using simple and
+- __small and consistent__: the library is about 6k LOC using simple and
   consistent data structures. This makes it very suitable
   to integrate and adapt in other projects. For runtime systems it
   provides hooks for a monotonic _heartbeat_ and deferred freeing (for
@@ -42,7 +42,7 @@ Notable aspects of the design include:
 - __first-class heaps__: efficiently create and use multiple heaps to allocate across different regions.
   A heap can be destroyed at once instead of deallocating each object separately.  
 - __bounded__: it does not suffer from _blowup_ \[1\], has bounded worst-case allocation
-  times (_wcat_), bounded space overhead (~0.2% meta-data, with at most 16.7% waste in allocation sizes),
+  times (_wcat_), bounded space overhead (~0.2% meta-data, with at most 12.5% waste in allocation sizes),
   and has no internal points of contention using only atomic operations.
 - __fast__: In our benchmarks (see [below](#performance)),
   _mimalloc_ always outperforms all other leading allocators (_jemalloc_, _tcmalloc_, _Hoard_, etc),
@@ -56,8 +56,9 @@ Enjoy!
 
 ### Releases
 
-* 2019-08-10, `v1.0.6`: pre-release 6: various performance improvements.
+* 2019-10-07, `v1.1.0`: stable release 1.1.
 * 2019-09-01, `v1.0.8`: pre-release 8: more robust windows dynamic overriding, initial huge page support.
+* 2019-08-10, `v1.0.6`: pre-release 6: various performance improvements.
 
 # Building
 

src/options.c

@@ -150,10 +150,10 @@ static _Atomic(uintptr_t) out_len;
 
 static void mi_out_buf(const char* msg) {
   if (msg==NULL) return;
+  if (mi_atomic_read_relaxed(&out_len)>=MAX_OUT_BUF) return;
   size_t n = strlen(msg);
   if (n==0) return;
-  // claim
-  if (mi_atomic_read_relaxed(&out_len)>=MAX_OUT_BUF) return;
+  // claim space
   uintptr_t start = mi_atomic_addu(&out_len, n);
   if (start >= MAX_OUT_BUF) return;
   // check bound
@@ -163,17 +163,17 @@ static void mi_out_buf(const char* msg) {
   memcpy(&out_buf[start], msg, n);
 }
 
-static void mi_out_buf_contents(mi_output_fun* out) {
+static void mi_out_buf_flush(mi_output_fun* out) {
   if (out==NULL) return;
-  // claim all 
+  // claim all (no more output will be added after this point)
   size_t count = mi_atomic_addu(&out_len, MAX_OUT_BUF);
-  // and output it
+  // and output the current contents
   if (count>MAX_OUT_BUF) count = MAX_OUT_BUF;
   out_buf[count] = 0;
   out(out_buf);
 }
 
-// The initial default output outputs to stderr and the delayed buffer.
+// The initial default output, outputs to stderr and the delayed output buffer.
 static void mi_out_buf_stderr(const char* msg) {
   mi_out_stderr(msg);
   mi_out_buf(msg);
@@ -195,13 +195,13 @@ static mi_output_fun* mi_out_get_default(void) {
 }
 
 void mi_register_output(mi_output_fun* out) mi_attr_noexcept {
-  mi_out_default = (out == NULL ? &mi_out_stderr : out);
-  if (out!=NULL) mi_out_buf_contents(out);
+  mi_out_default = (out == NULL ? &mi_out_stderr : out); // stop using the delayed output buffer
+  if (out!=NULL) mi_out_buf_flush(out);  // output the delayed output now
 }
 
 
 // --------------------------------------------------------
-// Messages
+// Messages, all end up calling `_mi_fputs`.
 // --------------------------------------------------------
 #define MAX_ERROR_COUNT (10)
 static volatile _Atomic(uintptr_t) error_count; // = 0;  // when MAX_ERROR_COUNT stop emitting errors and warnings